World Beta Blog

Lots of people commenting on a recent article from the very good blog Philosophical Economics written by @Jesse_Livermore.  I’m writing a book on the topic of global stock valuation models probably out in January (and having a design contest for the cover here now).  

An investor needs to be aware of the benefits, as well as the drawbacks of using any investment model.  Too many people follow their models and opinions with religious like zeal, much to the detriment of their portfolios.  Below we examine a few of the criticisms commonly heard when discussing the CAPE ratio.

 Measurement period is too long.  Critics claim recessions and expansions have an outsized impact long after they have faded from memory.   “Estimating Future Stock Market Returns” by Adam Butler and Mike Philbrick tackles the issue of different measurement periods from one year up to thirty (as well as other valuation models).   We take up this topic in an old blog post here and show the ideal period centers around seven years (thanks Ben Graham!).  But realize that 1 year PE ratio works pretty well too.  Note that CAPD works too!

It is impossible to compare across decades due to changes in accounting.   One of the things that makes me really queasy as a portfolio manager is when people make “adjustments” to historical price data.  The classic example is excluding 1987 from the analysis as an outlier.  Why would you exclude something that actually happened?  However, much like in our Shareholder Yield book, it is important to understand when there is a structural change in a market, and when is this time really different?

Critics complain about write downs and how that biases CAPE. However, critics also claim adjustments to CPI and accounting rules render comparisons across decades, or even centuries meaningless at worst, and at best inaccurate.  

I’m not an accountant, but we will examine some ways to think about the above criticisms and see if they impact the measure, and if so, by how much.  I really like Liz Ann Sonders at Schwab – she has a good piece on all things PE and makes a few comments:

 “More recently, the move toward fair-value accounting standards resulted in security losses having a devastating effect on the reported earnings of financial institutions during the recent financial crisis. Yet that effect now appears to have been transitory. If an accounting item is deemed non-recurring, it’s common practice to ignore it when determining underlying earnings (i.e., using “operating” instead of reported earnings). But CAPE continues to reflect the effect of non-recurring items for the 10 years that follow their initial recognition in reported earnings.”

The blog Philosophical Economics examines how reported earnings in the US have been inconsistent over time.  Shiller uses Generally Accepted Accounting Principles (GAAP) earnings from S&P, also known as reported earnings.  However, the early 2000s witnessed the introductions of FAS 142/144 which altered how companies amortize goodwill.  I will not bore you with a long overview, but this has the potential effect of biasing earnings down, and CAPE up.  Another CAPE critic, Jeremy Siegel, penned this note in the FT back in August 2013 in “Don’t put faith in CAPE crusaders”:

 “I believe the Cape ratio’s overly pessimistic predictions are based on biased earnings data. Changes in the accounting standards in the 1990s forced companies to charge large write-offs when assets they hold fall in price, but when assets rise in price they do not boost earnings unless the asset is sold. This change in earnings patterns is evident when comparing the cyclical behaviour of Standard and Poor’s earnings series with the after-tax profit series published in the National Income and Product Accounts (NIPA).”

Siegel has a white paper “The Shiller CAPE ratio: A New Look” and powerpoint presentation here.  (I added his charts at the end, but note all measures show some overvalucation currently.)

I used the series from Bloomberg mentioned in the Philosophical Economics article, and you can see the difference in the below Figure, and note the similar tracking until the early 2000s.

FIGURE :  EARNINGS COMPARISONS, 1954-2012

 Source:  GFD, Bloomberg, Philosophical Economics

Which series is “correct”?  Well, they both are, but perhaps the red line is more consistent across time.  I honestly don’t know, but a better question is, does it even matter?

Below is a Figure of both CAPEs, adjusted and non-adjusted.  While the CAPE declines from ~25 for the reported earnings series to ~21 for the non-GAAP series, both reach the same conclusion generally and currently, albeit with different magnitude .  US stocks are not cheap.  While we agree there may be some variation, in our paper we examine the CAPE in over 30 foreign markets with supporting results.

FIGURE :  CAPE COMPARISONS, 1970-2012

 Source:  GFD, Bloomberg, Philosophical Economics

A similar take on the topic is from Societe Generale who put out an excellent piece titled “To Ignore CAPE is to Deny Mean Reversion”.  They use the MSCI earnings index that doesn’t include the writedowns and they come to the same conclusions as using the S&P series – some overvaluation. 

Recessions bias CAPE up.  Bubbles bias CAPE down. People often find a way to justify their market stance.  I’ve humorously received both of these critcisms from market bulls and bears!  Here is a sample from one of my friends “We don’t like using the CAPE because it includes 2008-2009 earnings which distorts the PE since earnings are too low.” My response to this is, well, according to your logic, do you also exclude 1999 and 2007 as being abnormally high?  And then, if you make the adjustments, does it even matter?  This is a similar, but slightly different argument (one off recessions) than the prior one (an accounting inconsistency). 

Below we adjust the earnings series from Shiller to pretend like 2008/2009 never really happened.  We adjusted the earnings series so that earnings didn’t decline in 2008 and 2009 (they had already started to decline a bit in 2007).  The second chart is the adjusted CAPE series.  If you adjust the data it moves the CAPE from approximately 25 to 23.  There is basically no difference and stocks are still expensive, but not terribly so due to the mild inflation sweet spot we are in.

FIGURE :  ORIGINAL AND ADJUSTED EARNINGS, 2000-2013

 Source:  Shiller, GFD.

FIGURE 12:  ORIGINAL AND ADJUSTED CAPE, 2000-2013

 Source:  Shiller, GFD.

CAPE isn’t really a short term timing measure for one market.  Like most valuation measures, it is a blunt tool and it’s not that helpful telling you what to do for the next few months.  It makes much more sense to align the indicator with the measurement period. However, pretty much every value measure we track aligns to say the same thing – US stocks are expensive.

As Dr. Hussman shows in his weekly commentary and in his chart below, it doesn’t really matter which market valuation metric you prefer, most signal a bit of overvaluation to the market.  It’s nothing nearly as awful as the late 1990s, but it means that until this valuation “burns off”, which can take years, decades, or possibly even a month or two if we had a crash, we will have somewhat muted returns of perhaps 2-5% nominal per year.  Below are his charts that examine some basic valuation metrics.  (We also examined this in our paper on the FF data with totally different metrics and dataset with similar conclusions, value works!)

FIGURE :  VALUATION METRICS 1940 – 2012.

Even Mr Bogle uses a different metric – P/B on the Dow (from BI).

CAPE and other valuation methods are interesting on a stand-alone basis – but in this global age why focus on only one country?  Much more important is expanding the opportunity set to include all of the countries in the world and buying the cheapest instead of the owning the most expensive and often biggest by market cap. 

Siegel’s Charts:

Investors love chasing returns, always to their detriment.  The best place to be allocating likely is the worst performers over the past 3-5 years.  And if you have a strategy you believe in, the same.  Lots of redwoods crashing  down the last few years due to the below bloodshed, not a lot of performance fees to be had with three negative years in a row…

Source:  Red Rocks via Idea Farm.

Good to see Rob get a life time achievement award this past week in AZ, also good to see RA fundy assets cross $100B.

Here he chats with Tom Lydon, and near the end he talks about just how low the CAPE is on their fundamental emerging.  Preview:  it’s really low.

I moved The Idea Farm to a private list because it was getting too big, and many publishers didn’t want their research going out to a large audience, or available freely on the internet.  I have enjoyed sending out over 100 pieces of research in the past year, but am faced with a similar problem today as the list has grown.  Either a) close the list to new subscribers, or b) raise the price.  I don’t want to shut anyone out just yet, but I still want the list to be broadly affordable.  Beginning in 2014, I am going to raise the price.  I imagine if it gets much larger I will just close the list as I don’t really want to charge much more than this…

In any case, all current subscribers through 2013 will always pay the lower rate of $199/year, forever.  

I’m not sure why the Investors Intelligence (which polls advisors) is so much more bullish than AAII (individuals), maybe perhaps since individuals still haven’t bought into the market rally?  Curious to hear thoughts here:

I thought this would be a fun way to visualize the Dow stocks, and how they distribute their cash through dividends and buybacks.  Most friendly on the left to least friendly on the right. Note that some companies, like Cisco and J&J, seemingly have a good yield but are net issuers of stock…(Note this says nothing about valuation.)….

First list would be Dogs of the Dow (Dividend Yield)

Second list is (Cash) Cows of the Dow (Dividends & Buybacks, aka Net Payout Yield)

Patrick O’Shaughnessy has a great piece this month where they touch on a topic that is incredibly important now.  We mentioned this back in August where the premium that dividend yield stocks are trading at relative to the market is near the highest ever.  Historically when you invest in high yield you are getting a value tilt, but now, as money has rushed into all things high yield, you are actually getting the opposite – not something you want!

Valuations are also cheaper abroad.  

Conclusions?  

1.  In the US avoid high dividend yield in favor of shareholder yield.  

2.  Look abroad. (this chart from the summer).

I’ve been publishing CAPE updates for countries quarterly on The Idea Farm, and below I highlight a blurb from our upcoming year end outlook.  This chart shows the returns to country ETFs and the 10 cheapest and 10 most expensive markets.  Notice why I was so unpopular in Bogota in January when I said they have one of the most expensive markets in the world!  Also notice the big outlier in the expensive country bucket (the US). Due to all of the expensive countries declining and the US appreciating, we are now the most expensive in the world.

Also note the explosive returns in the cheap countries.  (Portugal only recently had an ETF launch).

As Rasheed Wallace would say, the ball don’t lie!

This post is similar to the recent post we did on F-Squared.  We sent out a research piece recently to The Idea Farm list from Pictet, a multi-billion $ asset manager out of Europe.  I am slightly embarrassed to admit I had never heard of them until recently when a reader emailed me some of their work.  They have one momentum strategy they describe as:

“One of our most original and historically successful approaches is our “momentum” strategy that allocates invested capital systematically between four asset classes. For those not familiar with this approach, it selects from US 10-year Treasury bonds, US equities, emerging-market equities and gold, and allocates 100% of the capital to the asset class that has shown the best performance in the recent past.”

That’s awesome!  You know I like it, although without a trend overlay and only selecting one asset can lead to some pretty wild swings.  Just how wild?  Below are the backtested results to 1973, 17% a year isn’t bad!

(4 asset classes updated monthly, ranked on 12 month total return.)  Granted if you used more asset classes and invested in the top 1/3 of assets your risk adjusted returns improve a bit…

Long time readers know that I am a big fan of simple rules based portfolios, heck that’s behind most of everything I do, from the buy and hold and 13F portfolios of The Ivy Portfolio to the trend portfolios of a QTAA, to shareholder yield approaches to income.  Frankly most all of the 2&20 world can be deconstructed for next to nothing.  For example, the book Following the Trend: Diversified Managed Futures Trading was actually really good – and I feel like it is pretty rare to say that these days.  It basically lays out how to replicate the vast majority of the managed futures industry with a simple system(s). (Covel’s book is great too.) It reminds me a bit of that Bridgewater piece on replicating basic hedge fund strategies with rules based investing  : Hedge Fund Returns Dominated by Beta – May 3, 2012

I was going to lay out a simple model, one very similar to the one we published back in 2010: Relative Strength Strategies for Investing.  This paper was a domestic expansion of work we published way back in 2007 in our Quant Approach to TAA.  

 I thought I would demonstrate the utility of another relative strength approach from F-Squared,  a $15b shop that a lot of RIAs use to outsource their tactical allocations.  (Note: this post is updated at F Squared’s request to only use their 2008 forward index data. )  

You can find their construction rules here:

• When fully invested, the model index all nine of the U.S. equity sectors:  At the point of rebalancing in a fully-activated mode, the strategy is equally weighted in each sector at 11.1%.
• The critical process, executed on a weekly cycle in the AlphaSector Premium Index, is the model’s review of each of the nine sectors to be either included or excluded from the portfolio based on likelihood of forward-looking positive return.
• The decisions are generated through a sophisticated analytical engine that evaluates “true” sector trends while adjusting for market noise and for changing levels of volatility in the market.
• The key model inputs (driving the decision-making process of the algorithm) are data on total return movements, volatility, and rate of change in volatility for the subject equity sector.
• The model output is a binary decision.  If a sector receives a positive signal for investing, it is included in the index portfolio.  If a sector receives a neutral or negative signal, it is removed.  All sectors represented are equal weighted, with a maximum allocation capped at 25% of the Index at the time of rebalancing.
• If there are three or fewer sectors represented at a given time, the remainder of the portfolio (reflecting the 25% maximum cap per sector) is invested in the Short-Term Treasury ETF, representing cash.  The Index can be 100% invested in the cash equivalent if all sectors receive a neutral or negative signal for investing.
• The presence of a cash equivalent position in the index portfolio during bear markets is a clear illustration of the F-Squared’s philosophy of “client-centric not benchmark-centric.”  Conventional U.S. Equity investment strategies would continue to track to the S&P 500 during a bear market, seeking to achieve only relative outperformance.  Investors are subject to potentially severe drawdowns, even while their traditional manager is perceived as “beating peers.” In contrast, the AlphaSector model breaks the correlation to the S&P by deploying the cash equivalent.  Delivering downside risk controls is our approach to meeting client needs.

My guess was that a simple system, similar to many we have published in our white papers, would capture what F2 is trying to do.  Below we examine 9 sectors,  equal weighted if above 10sma.  If less than 4 sectors then 25% in cash if 3, 50% cash if 2, 75% cash if 1.  We used the French Fama data that goes all the way back to the 1920s…The first chart is the test back to 2001 , the second chart is FF all the way back to the 1920s.

Note that the system does a good job of  capturing what F2 does through their public index.  Also note the strategy does a good job of reducing risk through vol and drawdown back to the 1920s.  Note those looking for outperformance should consider a more concentrated portfolio that only owns the top 2-4 sectors, and we will follow this piece up in a week with another killer momentum system published by another multi-B shop out of Europe…stay tuned!